Keynote at GX Connect 2010 (December 2nd), Amsterdam Arena, Amsterdam, The Netherlands.

1. Selling online
Analytics driven
content optimization
Ton Wesseling (@tonwesseling)
Online Dialogue (@onlinedialogue)
GX Connect 2010, December 2nd 2010, Amsterdam, The Netherlands

2. The partners
Ton
Bart
Anouschka
Maurice
Measure
Strategy
Design
Social

3. A network organization
To be flexible
and follow our passion
CC: http:/http://www.flickr.com/photos/marc_smith/3976381424

4. Our goal
Deliver valuable online experiences…
For the online user
CC: http://www.flickr.com/photos/soundfromwayout/143822346

5. Our goal
…while making money
for our clients
CC: http://www.flickr.com/photos/soundfromwayout/143822346

6. About Online Dialogue
 Online Dialogue has the ambition to get the possible quality of
offline service en sales dialogues to the online field.
Strategy
Design
Optimization
Our services:
 Internet strategy guidance.
 Designing the online presence.
 Getting structure in online revenue optimization.

7. Learning

8. Dialogue model

9. Dialogue processes
Theory: © http://onlinedialogue.com

10. What is online content?
Your part of the conversation!
32/60

11. What is (online) marketing?
Bringing in the people
and persuading them to buy

12. The Future
By profile targeting
and social recommendation:
“Good traffic is easy”
CC: http://www.flickr.com/photos/h-k-d/3629569854/

13. Just like 300 years ago
Product, Quality, Service & Price
Selling by conversations
26/60

14. Conversation skills will be the key
We invented ʻverbalʼ 100.000 years ago
35/60

15. Basic conversation rules
Donʼt just talk, listen!
CC: http://www.flickr.com/photos/steelight/70137572

16. Basic conversation rules
Respond to questions!
CC: http://www.flickr.com/photos/steelight/70137572

17. Basic conversation rules
Do not ask a question that you
would not want to answer yourself
CC: http://www.flickr.com/photos/steelight/70137572

18. Online markets are conversations
Are you doing this
with your online presence?

19. Conversation skills
Storytelling
35/60

20. 35/60

21. 555 years ago
35/60

22. 100 years ago
35/60

23. 80 years ago
35/60

24. From 1 to many
Masscommunication
28/60

27. The Web
connections
1-on-1
32/60

28. Everyone is connected
37/60

29. The web is like a small village…
36/60

30. dialogues
32/60

31. Dialogues…

32. Dialogues…

33. Dialogues…

34. Dialogues…

35. Dialogues?

36. Online behavior = clicks & typing
No verbal
No non-verbal
Information freedom
Time & place freedom

37. Company goals €€€  Customer goals
Company Prospect / client
Behavior
Behavior
Mul9 Mul9 Mul9‐
purpose phase channel
theory © http://onlinedialogue.com
icons © http://dryicons.com

38. Company goals €€€  Customer goals
Company Prospect / client
Behavior
Behavior
Mul9 Mul9 Mul9‐
purpose phase channel
External influencers
Compe9tors Media Social environment
theory © http://onlinedialogue.com
icons © http://dryicons.com
Tuning the conversation

39. Company goals €€€  Customer goals
Company Prospect / client
Time & Money Knowledge
Behavior
Behavior
Knowledge ACtude
Mul9 Mul9 Mul9‐ Personality
Tools
purpose phase channel
Internal
PlaHorm Profile
influencers
Internal Internal
Mo9va9on
influencers Context
influencers
External influencers
Compe9tors Media Social environment
theory © http://onlinedialogue.com
icons © http://dryicons.com
Based on internal and external influencers

40. Build up groups of customer profiles
segmented on their behavior
and collect the knowledge of
how to communicate with them

41. Very large 1:1 crowd
in groups
CC: http://www.flickr.com/photos/michaelroper/2618105/

42. You can live your own Groundhog Day

43. You just have to test
Itʼs not what youʼve said
Itʼs the visitors perception!

44. You just have to test
What will they choose?

45. To know = to research & measure

46. But how?

47. Are you ready?
Targets
Trends
Facts
Survey
Roadmap
Survey
Survey
1 – to get things done

48. Your audience as a group
Survey

49. Your audience as a group
Company
Target
info
audience
Revenue
Happiness
External afcts
Internal facts
Customers
Wealth

50. Your audience as a group
External survey
Measurement
Price position
Media attention
Attention
Research
Quality position
Advertising pressure
Positive / negative

51. Your audience as a group
Company
Target
info
audience
Survey
Revenue
Hapiness
External afcts
Internal facts
Customers
Wealth
External survey
Measurement
Price position
Media attention
Attention
Research
Quality position
Advertising pressure
Positive / negative
2 – to know your position

52. Make a good start
Time to (re-)design.. if needed

53. Dialogue processes
Theory: © http://onlinedialogue.com

54. Your audience in control
3 – to know if they succeed
Spontaneous
Orientating
Comparing
Buying
Cust. info
Cust. usage
Cust. slae
Targets
Growth
Trend
Conversion

55. Continuous improvement
Hypot
hesis
Alternative
Optimization process

56. Continuous improvement
Default
Variation
A
Variation
B
Winner!
Optimization technique

57. Continuous improvement
Structuring
A) Analysis
B) Roadmap & Dashboard
-
Conversion op9miza9on
1) Mass Usability
2) Segments
3) Dynamic
4) 1‐on‐1

58. Dashboard example

59. internal search
optimization with
instant search
Test Case

60. Index
1. Case background
2. What did we test?
3. The results!

61. Case background
Dutch Ecommerce
company Fonq is
the owner of the
fonq.nl online shop
and more than 60
niche shops that
are based on the
same lay-out as
the fonq.nl shop.
Fonq belongs to
the top 100* Dutch
Qweery is een Nederlands bedrijf dat zoek- en navigatieoplossingen biedt sites
E-business
voor E-commerce websites. Zij kunnen de aanwezige informatiebased on turnover
logisch
bundelen en op de juiste plek in de juiste vorm aanbieden. estimates. (*Twinkle)
In deze case beschrijving specifiek aandacht voor de interne zoekmachine met
filters oplossing en de Frequently Asked Questions (AJAX techniek) module.

62. Case background
Every shop of Fonq has an internal search module in the top navigation of the
website. We already knew that people using the internal search are very valuable
when it comes to turnover per visitor. The Fonq web stats told us that the internal
search users have a 250% higher conversion rate compared to the average user.

63. Case background
step 1
step 2
Once you enter a
search term or
sentence in the input
box of the current
situation, youʼd have
to submit the query
before the website
would show the
results on the next
page. We thought this
process was not
efficient enough!

64. Case background
New technologies over the past years gave us instant feedback opportunities at web
forms and auto complete functionalities in the web browser or even the search
engine. This all is saving us time and helping us to enter the right search query.
According to Hitwise, only 75% of al first queries to a search engine result in a visit to
a result from the result page. This “success rate” at internal search engines from web
shops even lowers all the way down to less than 50% (Online Dialogue research).

65. Index
1. Case background
2. What did we test?
3. The results!

66. What did we test?
We implemented an
insant search solution
as a proof-of-concept
test on 5 of the Fonq
webshops
- Skottelbraaishop.nl
- Curverwebshop.nl
- Brabantia-webshop.nl
- Barbecueswinkel.nl
- Fonq.nl

67. What did we test?
This is what the
implementation
looked like: every
new character input
would give you an in-
page search result
update with picture,
text and price on
each product.
The layer already
shows up after typing
the first character and
withholds only the
products from the
web shop the user is
visiting at that
moment.

68. Index
1. Case background
2. What did we test?
3. The results!

69. The results!
Default (version A)
TypeResult (version B)
6722 unique users
4095 unique users
318 conversions
289 conversions
Conversion rate = 4,73%
Conversion rate = 7,03%
This test was performed in 3 full weeks. Starting: May 17th 2010 en ending: until June 6th 2010

70. The results!
Internal search users
conversion rate uplift:
49%
This is what the
implementation
looked like: every
new character input
✓ Results are completely statistical significant*
would give you an in-
*More on this at: http://en.wikipedia.org/wiki/Statistical_significance
page search result
update with picture,
Extra yearly turnover byprice ontext and internal
search improvement for the 5 testedeach product.
web shops:
The layer shows up
€ 329.198
after typing the first
character and
presents only the
Prediction based on weekly gain from the web
articles during the test
shop the user visits.

71. Continuous improvement
Hypot
hesis
Alternative
Optimization process

72. Index
Valk.com case
Sorry, in Dutch 

73. Valk.com case

74. De 3 hypothesen
De pagina is Extra promotie van de We moeten de klant
onduidelijk: er moet last-minutes, overtuigen bij ons te
een duidelijke aanbiedingen en boeken en dat doen
hotelkamer boek arrangementen leidt we door het gratis
actieknop in beeld, dit tot meer unieke kopje koffie en de
leidt tot meer bezoekers die iets gratis WiFi onder de
conversie.
kopen.
aandacht te brengen.

75. De variaties: nieuwe basis variant
Scanbare tekst met daarbij het
benadrukken van de enige
juiste koopargumenten: snel,
veilig en 24 uur per dag!

76. De variaties: variant B
Alle 3 de hypotheses verwerkt in 1
pagina: kofie en wifi, aabiedingen,
lastminutes en arrangementen en een
duidelijke boek hotelkamer knop.

77. De variaties: variant C
Om te testen wat nu het effect is van
de koffie en WiFi aanbieding, dezelfde
variant als B, maar dan zonder deze
aanbieding!

78. De variaties: variant D
In plaats van arrangementen, last
minutes en aanbiedingen nu weer een
koffie en WiFi variant om te kunnen
bepalen wat de effecten van beide
opties zijn.

79. De variaties: variant E
Tot slot de uitgebreide variant van D om
ook weer te kunnen vergelijken met B
en de impact van de boodschappen te
kunnen doorrekenen.

80. Welke variant was de winnaar?
A
B
C
D
E

81. Welke variant was de winnaar?
D
+ 13,6%

82. De uitkomsten: conversie uplift
Variant
Conversie uplift %
Zekerheid
A
default
91,5%
B
-6,1%
98,2%
C
0,1%
90,6%
D
13,6%
winnaar
E
-15,1%
99,9%
De testperiode was 6 t/m 16 februari 2010 op een deel van het bezoek. Nieuwe tekst uplift: 2%
Zekerheid is de kans dat de winnaar in de volgende test weer van deze variant wint.

83. De uitkomsten: hypotheses
De pagina is Extra promotie van de We moeten de klant
onduidelijk: er moet last-minutes, overtuigen bij ons te
een duidelijke aanbiedingen en boeken en dat doen
hotelkamer boek arrangementen leidt we door het gratis
actieknop in beeld, dit tot meer unieke kopje koffie en de
leidt tot meer bezoekers die iets gratis WiFi onder de
conversie.
kopen.
aandacht te brengen.
Waar!
Niet waar!
Geen verschil

84. Continuous improvement
Hypot
hesis
Alternative
Optimization process

85. The road to take

86. Just like they did…
26/60

87. and just like him…

88. Remember
Content = Conversation
Measure
Improve

89. Schrijf je nu in!*
*Sorry for this Dutch promo slide 
Deze case is gepresenteerd op de
Online Dialogue donderdag.
Je kunt deze 2 maandelijkse kennis
sessie in Utrecht sessie zelf bijwonen!
Schrijf je nu in:
http://onlinedialogue.com/donderdag
Volgende sessies:
•
13 januari 2011
•
10 maart 2011
•
12 mei 2011
•
8 september 2011
•
10 november 2011

90. Contact:
Ton Wesseling
www.onlinedialogue.com
info@onlinedialogue.com
Tel: +3130 4100 170
Subscribe to our free newsletter at our site!
@onlinedialogue @tonwesseling